1. Field of the Invention
The present invention relates to a vehicle air conditioner installed in a vehicle having a front seat and a rear seat.
2. Description of Related Art
One type of conventional vehicle air conditioner includes a front air conditioning unit for front seats of a vehicle and a rear air conditioning unit for rear seats of the vehicle. The vehicle air conditioner uses a linear model to determine target blowout temperatures of the conditioned air from the front air conditioning unit and the conditioned air from the rear air conditioning unit.
The linear model has air conditioning load factors as variables, which include a set temperature, an inside air temperature, an outside air temperature, and solar radiation intensity. The set temperature is set as a desired temperature in a passenger compartment of the vehicle by the driver and a passenger. The inside air temperature in the passenger compartment is detected by an inside air temperature sensor. The outside air temperature is detected by an outside air temperature sensor. The solar radiation intensity is detected by a solar radiation sensor. The vehicle air conditioner determines its blowout port modes and blowout air quantities based on the target blowout temperatures.
A vehicle having a front row and only one rear row of seats is fitted with another type of conventional vehicle air conditioner including a front air conditioning unit for the front seats and a rear air conditioning unit for the rear seats. In order to achieve more comfort, the vehicle air conditioner uses neural networks to determine the target blowout temperatures, the blowout port modes, and the blowout air quantities of the conditioned air from the front air conditioning unit and the conditioned air from the rear air conditioning unit. The neural networks are learned by training data in advance. Air conditioning load factors such as stated above are inputted into the neural networks, which output the target blowout temperatures, the blowout port modes, and the blowout air quantities. Therefore, it is under consideration that the air conditioner installed in a vehicle having three rows of seats or relatively many rear rows of seats should be controlled by means of neural networks to achieve more comfort.
However, control by means of neural networks requires a mass storage device (ROM), and accordingly it costs more than control by means of linear models. In particular, if the air conditioner installed in a different vehicle having relatively many rear seats controls them by means of neural networks, the rear air conditioning unit for the rear seats is accordingly complex in structure, or two or more rear air conditioning units are required for the rear seats, thereby greatly increasing the cost.